41st Lunar and Planetary Science Conference (2010) 2454.pdf

THICK ICE DEPOSITS IN DEUTERONILUS MENSAE, MARS: REGIONAL DISTRIBUTION FROM RADAR SOUNDING. J.J. Plaut1, J.W. Holt2, J. W. Head, III3, Y. Gim1, P. Choudhary2, D. M. Baker3, A. Kress3, and the SHARAD Team. 1Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, [email protected], 2University of Texas Institute for Geophysics, Jackson School of Geosciences, University of Texas, Austin, TX 78758, 3Department of Geological Sciences, Brown University, Box 1846, Providence, RI 02912.

Introduction: A class of landforms typically found the SHARAD observations, simulated radargrams de- in the mid-latitudes of Mars, including "lobate debris picting the position and expected intensity of echoes aprons" (LDA) and "lineated valley fill" (LVF), had from off-nadir surface topography ("clutter") were long been hypothesized to contain a substantial frac- generated and compared with the original SHARAD tion of water ice [1-6]. Observations by the SHAllow radargrams. In addition, potential subsurface returns RADar (SHARAD) on Mars Reconnaissance Orbiter were converted from a time-delay to a depth geometry (MRO) have confirmed that these features consist pre- assuming a wave propagation speed in ice, to evaluate dominantly of water ice in at least two regions of Mars, the relationship of the detected interfaces with sur- Eastern Hellas and Deuteronilus Mensae [7,8]. This rounding exposed topography [e.g., 7,8]. Finally, adja- finding is consistent with a formation mechanism in- cent or overlapping tracks were examined where avail- volving the retreat of formerly extensive ice sheets and able to check for consistency of the subsurface detec- cold-based glaciers, combined with development of a tions. Detections indicated in Figure 1 were considered protective surficial lag of debris shed from nearby high robust and worthy of mapping if they occurred in areas topography and/or exposed by sublimation [9-11]. In free of clutter, displayed the expected depth geometry this study, we report on the regional distribution of and were seen in multiple tracks where possible. these thick (100s of m) ice deposits in the Deuteronilus Discussion: Mapping of subsurface interfaces indi- Mensae area, based on a comprehensive mapping cates that thick masses of ice are quite common in the campaign by SHARAD on MRO. Deuteronilus Mensae area (Figure 1). Detections are Deuteronilus Mensae: Deuteronilus Mensae (40- widespread, at the bases of mesas and massifs, along 51° N, 14-35° E), part of the dichotomy boundary linear and curvilinear dichotomy boundary scarps, and “fretted terrain” [12] contains a high concentration of confined within valleys (LVF) or in some cases within LDAs and LVF [2-3] that occur at the bases of scarps impact craters. In many areas, the detection rate was of mesas, knobs, craters and valley walls [13]. Relief 100%; i.e., where the morphology suggested the pres- of the adjacent scarps is generally 1-2 km, and most of ence of LDA or LVF and the geometry was favorable the LDAs themselves have 300-800 m of relief relative for the observation, the reflector was indeed detected. to the surrounding valley floors. LDAs are typically For example, nearly every observation of the inner ~10 km wide, measured perpendicular to the trend of scarps of the quasi-circular (remnant impact crater?) the adjacent scarp, with a range of widths of 5-25 km features at 40N18E and 39N23E showed an unambi- [14]. Crater counting and cross-cutting relationships guous basal reflector. Similarly, in the area of numer- indicate that the surfaces of LDAs and LVF in this area ous mesas and knobs around 44N25E, almost all of the are mid-to-late Amazonian in age, while the surround- observations showed reflections at the expected loca- ing plains are Hesperian and most of the plateau sur- tions. While numerous detections were seen in the faces are Noachian [13,15]. LVF, the detection rate was sometimes lower than on SHARAD observations: SHARAD is an orbital the classic LDAs. For example, the filled valley near subsurface radar sounder on MRO operating at a center 41N35E showed a reflector estimated at >1 km depth frequency of 20 MHz. Vertical resolution is 15 m (free in one area, but did not show a reflector in other parts space), with a horizontal footprint of 0.3-1 km by 3-6 of the same valley. Some regional trends in detection km. As of December, 2009, SHARAD has obtained rate are observed. At the northern edge of the study over 250 observations of the Deuteronilus Mensae area area, where the knobs and mesas show less relief, some (Figure 1). SHARAD normally operates on the night- of the LDAs did not show the expected reflectors. A side of the MRO orbit, collecting data with ground- clear decrease in detection rate is seen in the eastern tracks trending from north-northeast to south- part of the area, where some large, classic LDA fea- southwest. Most observations to date were collected tures show little or no subsurface signature. Likewise, with this geometry, although some dayside tracks are detections in LVF seem to decrease toward the south- also shown in Figure 1. The mapping campaign has east. The decrease in detections could be caused by resulted in generally dense coverage, with only a few one or a combination of several factors: increased sig- gaps between adjacent tracks larger than 20 km. For all nal attenuation due to composition or internal struc- 41st Lunar and Planetary Science Conference (2010) 2454.pdf

ture; a thicker surface lag; roughness at the surface or the basal interface. The latter effect may be present References: [1] Carr and Schaber (1977) JGR 82, near 42N23E, where LDAs that developed on a rough 4039-4054. [2] Squyres (1978) Icarus 34, 600-613. [3] impact ejecta blanket do not show subsurface detec- Squyres (1979) JGR 84, 8087-8096. [4] Lucchitta tions, while the adjacent LDAs do. One interesting (1984) JGR 89, B409-B419. [5] Colaprete and Jakosky effect is observed in areas of highly transparent LDAs (1998) JGR 103, 5897-5909. [6] Carr (2001) JGR 106, or LVF: subsurface detections can appear in radar- 23571-23593. [7] Holt et al (2008) Science 322, 1235. grams from off-nadir buried interfaces. These reflec- [8] Plaut et al (2009) GRL 36, L02203. [9] Head et al tors are not predicted in the clutter simulations, nor are (2006) EPSL 241, 663. [10] Head et al (2006) GRL 33, they in the expected position relative to the nadir track. L08S03. [11] Morgan et al (2009) Icarus 202, 22-38. While these are among the very few cases where [12] Sharp (1973) JGR 78, 4073-4083. [13] Chuang SHARAD is sensitive "buried clutter," it is worth and Crown (2009) Geologic map of MTM 35337, checking for this effect in other study areas. The impli- 40337, and 45337 quadrangles, USGS. Sci. Invest. Ser. cations of the widespread presence of thick masses of Map. [14] Li et al (2005) Icarus 176, 382-394. [15] ice in the mid-latitudes of Mars are of course signifi- Kress and Head (2009) LPSC abstr. 1379. cant. This ice likely contains a record of climatic and other environmental conditions at the time of its depo- sition and flow. This may include evidence of habita- bility. The LDA and LVF deposits are a significant fraction of the known non-polar ice inventory on Mars, and they are intriguing targets for in situ exploration.

Figure 1. MOLA shaded relief map of the Deuteronilus Mensae area, with the groundtracks of SHARAD obser- vations shown as yellow lines. The red segments indicate positions of robust subsurface detections below LDA or LVF deposits.